 |
  |
MISSOURI DIVISION |
|
 |
 |
PROCEEDINGS OF THE SEPTEMBER 2000 POST EARTHQUAKE HIGHWAY RESPONSE AND RECOVERY SEMINAR HELD IN ST. LOUIS MISSOURI
INVENTORIES OF TRANSPORTATION NETWORKS BY STEVEN FRENCH
The Mid-America Earthquake Center is a consortium of seven universities. It's headquartered at the University of Illinois and includes MIT, Georgia Tech, Texas A & M, University of Memphis, St. Louis University, and Washington University here in St. Louis. It is a multiple disciplinary center focused on looking at the earthquake hazard in Mid-America.
This is an expected ground motion map for the area, centered, of course, on the New Madrid area. The red is .2 Gs, the orange is .15, and the yellow is .1 G. You can see it covers a fairly large area. This area has some very unique characteristics.
One of the unique characteristics of Mid-America is its named, the Heartland of America, The transportation system for the nation actually crosses this area of high seismic vulnerability. St. Louis and Memphis have an extensive road system and there are a lot in between crossing this seismically active area. The rail network spreads out of Chicago and crisscrosses the area. Of course we have the interstate system. All will be affected by seismic action. It's bad enough when we have earthquakes in California. When we get a big one in the Heartland, it will do a lot of damage to our transportation system and thus impact interstate commerce. The movement of people and goods will come to a sudden stop.
We have very low probability but high consequence events. We don't have the frequency of events to remind us and to bring it to the top of the political agenda. We have a very high consequence event that can do a lot of damage but the return periods are very long. The risk falls out of public awareness and off of the public agenda. We have a lot less awareness and a lot less mitigation that takes place because we have long periods between earthquakes, unlike California or even some other places like Utah and Pacific Northwest.
Seismic event effects in the Heartland will be very widespread because we have very different geology than California. The attenuation or the spread of the energy will be much more widespread due to the characteristics of our geology. In California it's not uncommon to have the effects attenuate over a ten or twelve-mile area. Here, because of the geology, we're going to feel the earthquake effects over a much larger area, over 100 miles. Our geology is very different than in California.
One of the problems with the earthquake hazard zone in Mid-America is political fragmentation. It's a piece of seven different states and it's not a majority of any one of those. It's about a quarter or an eighth of seven different states. So everybody has to think about the part of the state that has the problem. It's not widespread across the whole state so not everyone thinks about it.
We have a very vulnerable building stock and infrastructure systems that are more vulnerable than what you find in highly seismic areas like California or Japan. You have a transportation network that is more vulnerable. I suspect it is comparable to what you saw in Turkey.
I'm going to talk about the inventory of transportation at risk, a project that we put together, data sets we put together using GIS. We're trying to model the potential economic damage to the infrastructure network. It's useful to think about how Mid-America is different from California and from other potentially seismic areas.
One of the first things we did when we put together the Mid-America Earthquake Center was to decide that we were going to actually look at two projects: central facilities covering hospital, police, fire and shelters, primarily, schools and secondly, the transportation networks program.
I'm going to talk about the inventory of transportation networks and what we've done. We've assembled a set of GIS inventory. We got GPS units and collected primary data. There are a lot of data sets from Federal Highways and other sources that we put together to create an inventory of critical transportation facilities in Mid-America. We then used that data set to (1) analyze the characteristics of the inventory, (2) to try to understand what the problems are, and (3) to really guide our research in terms of transportation networks.
While I've listed highways, bridges, railways, waterways, ports, and airports, our primary focus has been highways, bridges, and railways. We looked at those as the key elements and the ones we put our primary focus on. We have not dealt with waterway transportation and airports as much as highway and rail.
The purpose of this project was to provide data for other research centers in the MAE Center and to provide data for users such as emergency managers. We set up a web base distribution of our data sets and are getting some use from the researchers. I must say were getting a little bit less usage than I would have expected from the emergency management community. That's probably due to the fact that the dissemination is usually at the end of the project. Hopefully, today, we can make people more aware of the resources that we put together.
We have completed inventories of the seven-state area. We have all federal-aid roads that include all U.S. routes as well as interstates in the seven-state area. The data primarily comes from the highway planning network data sets. We used the National Bridge Inventory to collect data on all of the highway bridges in the seven-state area.
The NBI is a good initial inventory of bridges. It is sorely lacking in some of the characteristics. We really need to model and to understand the seismic response on the bridges. It does not include enough seismic related information to really be useful in terms of modeling response to the bridges.
We're looking at ways to augment that but it requires a massive amount of fieldwork. We're actually creating a sort of priority structure to identify the bridges we think are most at risk to try and collect some primary data on those.
We have all the rail lines in the seven-state area from the Federal Highway Agency. We have rail bridges as well as the rail lines in the seven-state area. We have the navigable waterways from the Corps of Engineers, the ports and lock facilities, as well as airports and basic commercial networks in the seven-state area.
Most of this, as I said, is second data. We have simplified it, deleted some of the immaterial data that is not pertinent to earthquake issues. It´s a little bit easier to find things that are useful in the data sets.
Let's look a little bit at some of the characteristics of Mid-America inventory. If you look at the types of highways (rural interstate, urban freeway, rural arterial, or principal urban arterial) by miles and look at the pieces in the .2 G expected ground motion area, you'll see that principal arterials are the dominant number of miles in the network. If you look at the rural interstates, you actually get a sense that Arkansas, Illinois, and Missouri really have the lion's share of rural interstates. They have the most exposure in terms of interstate mileage to this .2 G maximum expected ground motion in the area. You'll notice that in Tennessee, we had a fair amount of urban freeway with Memphis. St. Louis will get a little bit of urban freeway and you also get some of the urban freeway within Arkansas in this high-risk area.
With the GIS, we can take this national data set and look at subsets and determine how much is in the expected ground motion areas. By building the data set the way we have, we're able to do those kinds of things and make those kinds of analysis.
Here's kind of a detailed look at the high-risk area. We have two and four-lane roads. You can see the point representation of the bridges where we have greater than 10,000-average daily traffic. We actually have a county representation of the high-risk areas.
The light gray is the .15 G and the darker gray is the .2 G area. You can see the Mississippi River coming down through there. You can actually see which bridges and which roadways are exposed to the most likely ground motion.
We look at the characteristics of the bridges. The bridges are really the most vulnerable parts of the roadway as we saw in the Turkey experience. We can get some general information but we really can't get bearing information or pier foundation information from the NBI that we'd like. We can get some general characteristics from the bridge inventory.
We have buildings going back to the turn of the century. We have quite a few built in the 1930s through WPA, and then quite a few built in the 1960s, 1970s, and 1980s. The distribution we have is fairly wide.
The good news is we certainly have a number of buildings that have been built to fairly modern standards. I think if you look at the standards in place in the 1960s and even the 1970s, I think you'd question whether those bridges are going to perform very well in the event of an earthquake. I certainly don't want to be on one of the 1900 bridges.
If you look at the structure type, we have a lot of concrete bridges in the high-risk area. That's the predominant structure type for most of the interstate and overpass facilities. There are also a significant number of steel but less in terms of prestressed concrete. And for the rest, you get sort of a smattering of other types such as wood, etc.
The superstructure condition is not really a condition assessed in terms of seismic response. It´s a condition in terms of just everyday functionality. If you look at the condition of the bridges, seven, eight, and nine are in good condition. But we have a fair amount in the five to six fair category. I don't know how good the correlation is going to be with the seismic response. But if it were only in fair condition, one would wonder if it's going to perform very well in terms of seismic response.
If you look at the maximum length of spans, you can actually see that we have a lot in the approximate 100-foot range. We have some with spans up to 400 feet and more. I assume the ones with the exceptionally long spans are the ones crossing the Mississippi of which there are only a few. Those are significant choke points in terms of transportation networks. Those are the characteristics of the kinds of spans we have to deal with.
The nice thing about the data sets is we have rectified them and set them up so that the bridges can be seen in association with the highways that they service. We're looking at Memphis and you can actually get in a fairly close level. If you look at the green bridges, those are in good condition. The Mississippi River is coming up on a gray blue, Memphis on the east and Arkansas on the west. The light blue, which are coming out a little bit on the purple side here, are in fair condition. The blue, which you see a whole string of along in Arkansas, are those in poor condition. So, again, these are not seismic inspections but I think it suggests the age of the bridge, the condition of the bridge, one would expect those to perform probably less well than newer bridges that are in better condition at this point.
This data set is then feeding into four subsequent projects that we're using. Our goal is to really understand how damage to the Mid-America transportation network will result in economic loss to the nation. We are modeling at this point interregional economic flows, looking at the flow of goods and services across sub-state areas in Mid-America and from areas external to Mid-America.
If those flows are damaged, we're looking at how long those flows will be interrupted and what kinds of economic losses will result. The economic losses may be as simple as the added transportation costs of just going around the damaged area. It may involve some business losses for people who can't get raw materials to their factory. The economic model looks at how interruptions to the transportation system will affect overall economic loss from disruptions to the network.
Two projects that we are starting in the next year are looking at the benefits and costs of different retrofit strategies. If we go in and apply mitigation measures to bridges, what does that cost? That's pretty straightforward. To determine the benefits are a little bit more difficult. That's the purpose of the network loss model that we can actually show by strengthening, or in some cases, creating more redundancy in the transportation system. We can have benefits and we can actually get some comparison of the benefits and the cost of those retrofit strategies.
We put that all together into an overall network vulnerability evaluation where we looked at a systems level approach to the whole transportation network to determine how vulnerable the whole system is, taking into account ultimate routes and redundancy in the network structure.
We've basically put together our base data. We are putting together the economic models that operate on top of this rather than traffic across this and having the ability to take out lengths and see how that changes the number flow.
We have both GIS and tabular data sets available on this web site for download. I see this data not being used just in a research context but in things like emergency planning and other kinds of uses.
One project we hope to do is integrate this with our central facilities data at a metro level to look at the accessibility of hospitals and other emergency service facilities after we damage the transportation system to integrate these two strands of our work.
 |
Send mail to scott.holder@fhwa.dot.gov with questions or comments about this web site. |
